• State TI Thinks Resolved
  • Locked Locked
  • Replies 15 replies
  • Answers 2 answers
  • Subscribers 34 subscribers
  • Views 398 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LMH0002: Output Level

user6308822
Prodigy 70 points
Part Number: LMH0002

Tool/software:

Hello, Team

We have a question about LMH0002SQ.
We use LMH0002SQE/NOPB for ASI driver.
In the specifications, the output (Vsdo) is written as 750 to 850mV (p-p) under the conditions (REF=750Ω1%).
When measured with the oscilloscope's maxhold setting, some ICs output more than 880mV.
This results exceed the specifications of the DVB-ASI standard.

We use this IC in a variety of products, all of which have the same circuit configuration.
Looking at some outputs, the results are low (819 (average), 832 (maxhold)) and high (869 (average), 884 (maxhold)).
There appears to be a difference in the output value depending on the LOT number. Also, there are some individuals with the same LOT number that are different.
The environment is the BNC connector at the DRIVER output circuit → BNC cable → 75Ω⇔50Ω converter → oscilloscope.
It says that the specified output (Vsdo) is 750 to 850mV (p-p) under the conditions (REF=750Ω1%), but does it vary as shown above?

Regarding the notation written in the data sheet, in terms of the oscilloscope, is it the result of max hold or the result of amplitude?
Under what conditions (measurement environment, circuit) are the results measured?

Best Regards,
Miyajima

  • 0
    TI__Genius 9209 points

    Hi Miyajima-san,

    Please allow me to check with my team and then I will get back to you shortly after.

    Best,

    Nick

  • 0 in reply to Nick Peabody
    TI__Genius 9209 points

    Hi Miyajima-san,

    The results in the datasheet are measured with a 1 meter BNC cable at the driver output of the LMH0002. The amplitude is then measured either with a video analyzer with a 75Ω input, or an oscilloscope with a 75Ω plug-in. In our measurement setup we do not use a 75Ω⇔50Ω converter. 

    Best,

    Nick

  • 0 in reply to Nick Peabody
    Prodigy 70 points

    Hello, Nick

    Thank you for your reply.


    In your measurement environment, what GHz band do you use an oscilloscope for measurements?

    I have attached our measured waveforms.
    We measured with a 2GHz band oscilloscope, and the results are the amplitude and p-p (maxhold).
    As a result, the waveform contains noise. The amplitude does not exceed the DVBASI standard of 880mv, but the p-p (maxhold) slightly exceeds it.
    Does the output of this device include noise like the attached waveform?
    We will now confirm whether the problem is caused by our circuit.

    There are differences in the output depending on the LOT number and individual (approximately 50mV as far as I've seen a few), but does this tend to be biased depending on the LOT?

    Best Regards,
    Miyajima

  • 0 in reply to user6308822
    TI__Guru**** 176281 points

    Hi Miyajima-san,

    Due to the US Labor Day holiday, the response to your question may be delayed. We will look into your question when we return tomorrow, Sept 3rd. Sorry for the wait and any inconvenience it may cause.

    Thanks,

    David

  • 0 in reply to user6308822
    TI__Genius 9209 points

    Hi Miyajima-san,

    We use a 23 Ghz bandwidth scope plug-in for scope measurements. We use it for the evaluation of 11.88Gbps SDI signals. According to the Nyquist theorem, 2GHz is sufficient to sample 1.485 Gbps signals.

    user6308822 said:
    There are differences in the output depending on the LOT number and individual (approximately 50mV as far as I've seen a few), but does this tend to be biased depending on the LOT?

    The datasheet variation of 750 mV to 850 mV accounts for the differences in part performance across LOT number. It is expected that you may see 50mV or more of a difference within the data sheet specified range.

    Best Regards,

    Nick

  • 0 in reply to Nick Peabody
    Prodigy 70 points

    Hello, Nick

    Thank you for your reply.
    Which output specification (p-p:750-850mV) is the result of maxhold or amplitude?

    Best Regards,
    Miyajima

  • 0 in reply to user6308822
    TI__Genius 9209 points

    Hi Miyajima-San,

    When using an oscilloscope, our measurements are based on the amplitude of the eye. To get the amplitude we would measure the distance between the peaks of the histogram.

    However, for the recent datasheet validation of our SDI parts we used a Phabrix video analyzer for amplitude measurements. See the image below from Phabrix that shows the amplitude on the bottom left.:

    The maxhold measurement is still relevant and is present in overshoot and undershoot specs. See the table below for more SMPTE specifications, specifically overshoot and undershoot.

    Best Regards,

    Nick

  • 0 in reply to Nick Peabody
    Prodigy 70 points

    Thank you for your reply.

    I understand that the specifications in the datasheet are amplitude, not maxhold.

    Sorry to ask about the details."
    It says "Supports DVB-ASI at 270Mbps", but the DVB-ASI standard(EN 50083-9:2002) says the "OutPut Voltage" is 880mV (p-p), so there seems to be no regulations such as overshoot or undershoot.
    This device supports DVB-ASI, but strictly speaking, I understand that overshoot and undershoot are not taken into account in terms of DC specs.
    Is this correct?
    I apologize if my understanding is incorrect.

    Best Regards,
    Miyajima

  • 0 in reply to user6308822
    TI__Genius 9209 points

    Hi Miyajima-San,

    Thank you for sharing this document.

    user6308822 said:
    but strictly speaking, I understand that overshoot and undershoot are not taken into account in terms of DC specs.

    I'm sorry, can you re-iterate this question? I am a bit confused by what you mean by DC specs.

    I sourced this definition of overshoot from tek for our reference:

    "Overshoot within the eye display is likely due to reflection along the signal path and can be due to incorrect terminations. The user will have to test the signal at various points in the chain to determine from where the reflections are occurring, this could occur at a patch panel, barrel connector or problem with a cable or connections."

    Best Regards,

    Nick

  • 0 in reply to Nick Peabody
    Prodigy 70 points

    Hello Nick

    Sorry for the confusing question.
    According to our current output measurement results of this device, the maxhold is over 880mV, and the amplitude is within 880mV.
    Therefore, according to EN 50083-9:2002, the input specifications for DVB-ASI are stated to be within 880mV.
    I guess that the specifications of this device are 750-850mV at amplitude, which exceeds ASI's input specifications at maxhold.
    Strictly speaking, you can see that the flat areas above and below the swing voltage are mixed with noise from our measurement results.
    When measured with maxhold, it may exceed 880mV due to noise.
    Therefore, if you are considering this device based on overshoot and undershoot(10%=80mV), it is possible to output up to min=640mV and max=960mV, which would exceed ASI's input specifications.
    According to ASI's specifications, if the amplitude is 800mV+-10%, there is no problem, but since I have not found any such issue, I am thinking that the measurement should be performed at maxhold.
    What is the basis for your support for DVB-ASI?
    If you only consider the numbers, I feel that it exceeds the DVB-ASI specifications.
    You think that the input specs for DVB-ASI should be within 880mV for amplitude, so does this device support DVB-ASI?
    Sorry to ask about the details.

    Best Regards,
    Miyajima

  • 0 in reply to user6308822
    TI__Genius 9209 points

    Hi Miyajima-San,

    It is possible to exceed the spec of DVB-ASI with the LMH0002, however with proper device tuning, the DVB-ASI spec can be met. As you have said, if the amplitude is 800mV +-10% there is no problem. 

    Additionally, lower data rates tend to have a lower than 5% overshoot percent. What is the typical overshoot percent you are seeing? With our recent devices the overshoot percent for SD data rates can be as low as less than 2%

    Best Regards,

    Nick

  • 0 in reply to Nick Peabody
    Prodigy 70 points

    Hello Nick,

    Is there somewhere that the input standard for DVB-ASI is written as 800mV+-10% in amplitude instead of maxhold?
    I would like to know the evidence that there is no problem in making judgments based on the amplitude measurement results.

    My measurements show no overshoot or undershoot.
    However, as I said in the previous thread, it is noisy in the flat part and exceeds 880mV at maxhold, so I wanted to know the premise of the specifications of this device.
    I understand the premise because I heard it in a previous thread.

    Best Regards,
    Miyajima

  • 0 in reply to user6308822
    TI__Genius 9209 points

    Hi Miyajima-San,

    Apologies if my answer was unsatisfactory. I am not as familiar with the DVB-ASI standards.

    user6308822 said:
    My measurements show no overshoot or undershoot.
    However, as I said in the previous thread, it is noisy in the flat part and exceeds 880mV at maxhold, so I wanted to know the premise of the specifications of this device.

    To fix this problem, I still recommend that you review your test setup. The 75Ω⇔50Ω converter is causing reflections that may be causing the noise which would effect your maxhold.

    I see from your screenshot that the maxhold is 884mV. This is around 15 mV higher than the signal's mean of 869mV.  15mV/869mV *100 = ~ 2% increase. Based on what you've shared with me about the DVB-ASI spec, I recommend keeping the device mean amplitude lower than 860mV because 860 * .02 = 17, 860 +17 = 877mV, still under spec.

    Best,

    Nick

  • 0 in reply to Nick Peabody
    Prodigy 70 points

    Hello Nick,

    Thank you for your reply.
    I understand that the fact that this device is said to be compatible with DVB-ASI means that it has been confirmed that there are no problems with DVB-ASI, but you do not know much about DVB-ASI.
    Does this mean that it is unclear whether it complies with ASI standards?

    Thank you for your suggestion.
    However, this time we just happened to have an individual with a high typ out, and in the future there may be one with a low typ out.
    Therefore, I do not want to manipulate the swing voltage with the Ref resistor.

    Best Regards,
    Miyajima

  • 0 in reply to user6308822
    TI__Genius 9209 points

    Hi Miyajima-san,

    Based on the datasheet, we are clear that the LMH0002 can comply to the ASI standards, however, with your measurement setup it is unclear whether or not the device will meet the p2p ASI specification. 

    Best Regards,

    Nick